Renal cell carcinoma with metastasis to the pancreas: Genomic signatures and clinical outcomes.

Abstract

e16519 Background: Renal cell carcinoma (RCC) is a diverse disease characterized by several subtypes, mutations, and a broad spectrum of sites of metastasis. Previous publications have established that RCC with pancreatic metastases (PM), a rare subpopulation, has more favorable outcomes, even when other metastases sites are involved. Outcomes in this subpopulation tend to be independent of the traditional IMDC risk stratification. We have previously explored the genomic landscape of RCC with PM. In this analysis, we explore the clinical data and correlation with NGS for a subset of patients treated at our institution. Methods: We identified patients (pts) at our institution with metastatic RCC with PM. Included pts had pathologically confirmed RCC with met to pancreas seen on imaging. All identified pts had NGS performed with Tempus xT assay (DNA-seq of 648 genes at 500x coverage). We also extracted demographic, clinicopathologic, and treatment data for included pts through chart review. We describe baseline characteristics, most prevalent gene mutations on xT and overall survival (OS) from time of diagnosis to death with particular attention of OS stratified by PBRM1 mutation. We also performed a log rank test to compare OS for those with and without PBRM1 mutation. Results: 23 pts with metastatic RCC with PM were identified. Pts included were diagnosed with RCC between 1993 and 2020, median age was 62 years (range 42-78), 48% were female, 70% of white race, 35% with smoking history. All patients had a nephrectomy and 91% had clear cell histology. 6 pts (26%) had de novo metastatic disease (3 [14%] with de novo PM). Median time to metastatic disease was 3.5 years (range 0-15.5 years) and median time to PM was 4.7 years (range 0-15.5 years). The most prevalent genes identified with mutations were VHL (70%), PBRM1 (39%), and SETD2 (32%); TP53 mutations were identified in 2 pts (9%). After a median follow-up of 16 years, 10 pts (43%) were deceased (7 due to RCC, 3 due to other causes), 11 pts were alive (5 without disease and 6 with active disease) and 2 with unknown vital status. Median [m]OS for the whole cohort was 12.7 years, which when stratified by PBRM1 mutation was 8.4 and 12.7 years for those with and without PBRM1 mutation, respectively, though this difference was not statistically significant. Conclusions: In our cohort, patients with RCC and PM had an indolent course with a prolonged survival even after diagnosis of metastatic disease. This effect was independent of PBRM1 mutations or other common mutations. As these patients presented with both de novo PM as well as later in the course, this may represent a unique population, rather than a result of survival bias.